Tracking graph

[LewisM]

Below is a tracking graph screenshot. It was produced in PHD. The mount is a Vixen GPD2 with the Skywatcher Synscan GOTO add-on, the guidecam is a Lodestar mono, the guidescope is a converted 9x50 finder, and the scope was aTakahashi FC-100 reduced to f/6.4. Total rig weight = 9.1kg (nearly at mount published limit). CCD is an ancient SBIG ST-2000XM, with Baader 6nm H-a filter.

Polar alignment was done via rough compass and inclinometer, and then I used iterative 2 star alignments (Sirius and Arcturus - well spaced). I then adjusted the AZ and ALT using the mount bolts, and reran the iterations until the handset reported errors were sub-minute (the elevation came in at a reported 4 secs off, the az 11 seconds - of course that all depends on my precision using the crosshair overlay in the MaxIM short exposure result).

I cannot use the polar scope in my backyard (a tree blocks SCP).

Software used was MaxIM DL and PHD. MaxIM refused to guide, telling me the star was not in all frames (?), but PHD locked in without issue. PHD reported merely GUIDING most of the time, only issuing a guide pulse every now and then. Guiding exposure was set at 0.5 sec - the PERFECT seeing permitted this! (there was ZERO star twinkle that night, and I could naked eye observe MANY targets that night from the middle of suburbia)

So, that's the background. I was initially doing 10 minute exposures, but in the end went for 20 minute exposures, concluding with a trial of 30 minute single exposures. The target was IC2944 (here binned 2x2 for size), and was more a test of the system than a serious effort.

The image is 2x10 H-a combined.

Looking at the tracking graph, does it indicate a slight polar misalignment still, or would the slight roughness indicate just PE? I had re-adjusted the worm meshing just that day. Would the peaks indicate areas that are too tight, highlighting the minor imperfections in the spur gears?

Regardless, I am fairly satisfied with the result. Ideally, I need to go back to an 8300 chipped cam to get better sampling. I want to refine the PA, as the mount will be permanently situated in this spot (though not STRICTLY permanent, as it is on a half-pillar on the tripod, all weighted down further)

[DavidTrap]

Looks good, although another test would be trying your guide camera through the main scope and see what the graph looks like then. The little finder guider has fairly low magnification.

DT

[Shiraz]

that's looking pretty good Lewis :).

At your guider image scale, an RMS of 0.1 pixels corresponds to about 0.9 arcseconds RMS, which is OK. With the min motion set to 0.15, corrections will not be applied until the error reaches 1.35 arcseconds, so it would not have been called on to correct very often in those conditions. You might get even better guiding if you use a smaller value for this parameter. If you really want to check your guiding, use phd2 and enter your guider fl and pixel size - it will convert to arcsec and give you a guide graph that is more readily compared to others (and shows how much wander there really is - at your scale, the wiggles in the phd graph are difficult to make out). It will also show guide error RMS for each axis, which gives some idea of how much worm error there is in the RA.

However, the mount is definitely working well enough that the stars look nice and round at your imaging scale of 2.35 arcseconds (or 4.7 arcseconds binned), so you currently have a well balanced wider field system - the 4000 is probably a good choice with your scope and mount and the results look really good. Please, let's see more - sky permitting of course.

[LewisM]

Thanks David and Ray.

Ray, I would use PHD2 - and have it installed - but it just crashes continually on my system (Win 7 64 bit). I cannot seem to find out the reason why. It either freezes up or simply won't un-grey the guide button. Odd. So, I use the classic version.

MaxIM is usually flawless guiding for me, but it continually gave me the "Star was not in all frames" message, which is perplexing. I even used the large bright star in a guiding trial to try to resolve that after giving up on trying with the smaller stars. Still same result. I tried exposures from 0.5 sec to 3 sec in MaxIM (with new darks for each). No go. So I went with PHD again.


Will lower the minimum motion a touch and see if things improve a little more. HOPEFULLY tonight if this wind goes away (I am fairly protected though - I image behind our 3 story apartment)

Thanks. Good to be imaging again after 6 moths!

[AndrewJ]

Gday Lewis

Another option if you just want to test for polar alignment and drive errors is to use PEMPro. It has a 60day trial period and has inbuilt drift wizards and FFT analysis that can give good data of whats going on gear wise.

Andrew

[LewisM]

I did, but the trial expired and no matter what I clear in cache, registry etc it won't reinstall and let me trial it again. There must be a hidden sys file or name file somewhere (like what Pixinsight uses).

I downloaded a free PE analysis tool (in Czech, but I should figure it out!) and may try that for a while.

[AndrewJ]

Gday Lewis

Another option is Metaguide then.
In "analyse mode", it has extensive logging and can use video camera feeds as well.
It also creates logs formatted for use in PEC Prep, but you can also analyse the data in more detail fairly easily yourself if reqd.
I wrote an app to do FFT on its logs and its not difficult, as all the data ( and more ) is stored in its logs.

Andrew

[Shiraz]

the other thing to try would be to load your phd log into PECprep. that will show you the tracking error in arcsecs if you load your guide scope fl and pixel size - I don't think you should check the DEC offset correction box. It will show true tracking error and, although it won't show true PE, the curves it produces as "PE" should show how much PE is leaking into your guiding.

You can also use phd to check how good your PA is by just turning off the mount correction and seeing how much drift you get.